It has been long held that aging-related brain changes contribute to late onset depression (LOD). We argue that white matter microstructural abnormalities and abnormal activation in the emotional and the cognitive control territories contribute to the development of LOD and to the persistence of its symptoms. To our knowledge, the proposed translational study would be the first to focus on mechanisms of LOD using advanced multimodal neuroimaging methods. Our focus on control networks is based on the following observations: 1) Susceptibility to interference from negatively-valenced irrelevant stimuli (emotional control) may be particularly salient in depression; 2) Difficulty engaging in goal-directed behavior while ignoring irrelevant stimuli (cognitive control) is a cardinal feature f depression and cognitive control processes are preferentially susceptible to advancing age; 3) Poor performance on a traditional neuropsychological measure of cognitive control is associated with persistence of depression despite antidepressant treatment. Consistent with the NIMH RDoC Project mandate to identify clinically relevant models of circuitry- behavior relationships, the primary aim of this project is to use functional magnetic resonance imaging (fMRI) and diffusion tensor imaging (DTI) in individuals with LOD to examine whether dysfunction of the emotional and cognitive control systems are mechanisms by which aging-related brain abnormalities contribute to LOD. For our primary hypotheses we will focus on the role of: 1) Microstructural white matter (WM) abnormalities (measured by probabilistic tractography) and activation of emotional control structures in the development of LOD; and 2) Microstructural WM abnormalities and activation of cognitive control structures in the development of LOD. For our secondary hypotheses, we will focus on the role of microstructural WM abnormalities, activation of control structures and the persistence of LOD despite treatment with an SSRI. The aims will be pursued in 70 patients with LOD and 70 non-depressed older adults. The LOD group will undergo a 2-week, single-blind, placebo lead-in, psychotropic washout phase at the end of which they will complete an MRI session that includes fMRI and DTI. Patients will then receive 12 weeks of escitalopram treatment at the target daily dose of 20 mg. In addition to testing our hypotheses, the study enables exploratory analyses of: 1) The relationships among WM hyperintensities, control system activations, and the development and persistence of LOD; and 2) Relationships of control network activation post-treatment to a. baseline activation and WM integrity and b. persistence of depression at 12 weeks. We expect this MRI study of LOD to offer novel information about the neural circuitry mechanisms of depression. In addition, the identification of structural and functional impairments of LOD can aid the development of clinical instruments and targeted interventions.